Better Understanding, Better Fixes? A Study of Hallucination in LLM-based Automated Program Repair

• URL: http://arxiv.org/abs/2609.04909v1
• Date: Fri, 04 Sep 2026 09:11:27 GMT
• Title: Better Understanding, Better Fixes? A Study of Hallucination in LLM-based Automated Program Repair
• Abstract summary: Large language models (LLMs) have significantly advanced automated program repair (APR)<n>In APR, hallucination may arise not only in final patches but also in the intermediate artifacts that guide patch generation.<n>We characterize hallucination as the production of patches or intermediate artifacts that are not faithfully grounded in the available repair evidence.
• Score: 27.262974988013728
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large language models (LLMs) have significantly advanced automated program repair (APR), yet existing evaluations remain largely result-centric and provide limited insight into hallucination during repair. In APR, hallucination may arise not only in final patches but also in the intermediate artifacts that guide patch generation. To address this gap, we perform a multi-layered analysis of hallucination throughout the APR process. Specifically, we characterize hallucination as the production of patches or intermediate artifacts that are not faithfully grounded in the available repair evidence. We examine repair hallucination in final patches and understanding hallucination in intermediate artifacts through three tasks, namely triggering testcase identification, line coverage prediction, and additional testcase generation.We then evaluate three representative LLMs on 832 Defects4J bugs through automatic evaluation and manual analysis. Our results show that both repair and understanding hallucinations remain prevalent. Across models and settings, only 21.0%-55.9% of generated patches pass the developer-written test suite. Moreover, although more accurate intermediate artifacts are generally associated with successful repairs, this relationship does not always hold. Manual analysis of 812 sampled repairs identifies repair hallucinations in 72.7% of cases, including patches that pass all available tests; incorrect causal localization and incorrect repair strategies account for 45.9% and 18.5% of these hallucinations, respectively. Meanwhile, models frequently misidentify triggering testcases, mispredict line coverage involving branching control flow, and generate additional testcases with missing bug-triggering conditions or incorrect expected behavior.

Related papers

• Validation Evidence in LLM Repair Agents: How Much of What Passes Actually Tests the Bug? [1.6787220460106658]
  BSG-VA captures each validation command at its exact working-tree state, extracts a test-only patch, and replays the command on the original buggy code.<n>The captured outcome and the replay results assign every event an evidence role, from gold-aligned bug-discriminating through regression-only to misleading.<n>A three-arm experiment tests whether returning the B-replay outcome to the agent changes this pattern.
  arXiv  Detail & Related papers  (2026-07-30T22:32:14Z)
• Beyond Fail-to-Pass: Iterative Hardening of Co-Generated Bug Reproduction Tests and Fixes [55.114939648705395]
  Large language models (LLMs) have made automated program repair increasingly practical for real-world bugs.<n>Bug reproduction tests (BRTs) help close this gap by turning a bug report into an executable, bug-specific signal.<n>We propose CoHarden, a co-generation framework that uses the Lax signal as an in-loop convergence criterion.
  arXiv  Detail & Related papers  (2026-07-22T07:30:07Z)
• Empirical Analysis and Detection of Hallucinations in LLM-Generated Bug Report Summaries [0.7666240799116113]
  Large Language Models (LLMs) are increasingly used to generate summaries of software bug reports.<n>LLMs produce hallucinations that can be convincing but unsupported by the source report.<n>This can mislead developers and reduce trust in automated maintenance tools.
  arXiv  Detail & Related papers  (2026-05-22T18:55:46Z)
• Historian: Reducing Manual Validation in APR Benchmarking via Evidence-Based Assessment [0.19853810231896352]
  We present Historian, a framework that leverages Large Language Models to perform multi-reference comparisons against a knowledge base of historically validated patches.<n>In leave-one-tool-out evaluation, Historian achieves 95.0% coverage with 88.4% accuracy, reducing manual validation to 5% of patches.
  arXiv  Detail & Related papers  (2026-02-28T13:41:29Z)
• Counterfactual Probing for Hallucination Detection and Mitigation in Large Language Models [0.0]
  We propose Counterfactual Probing, a novel approach for detecting and mitigating hallucinations in large language models.<n>Our method dynamically generates counterfactual statements that appear plausible but contain subtle factual errors, then evaluates the model's sensitivity to these perturbations.
  arXiv  Detail & Related papers  (2025-08-03T17:29:48Z)
• Generate, but Verify: Reducing Hallucination in Vision-Language Models with Retrospective Resampling [78.78822033285938]
  Vision-Language Models (VLMs) excel at visual understanding but often suffer from visual hallucinations.<n>In this work, we introduce REVERSE, a unified framework that integrates hallucination-aware training with on-the-fly self-verification.
  arXiv  Detail & Related papers  (2025-04-17T17:59:22Z)
• The HalluRAG Dataset: Detecting Closed-Domain Hallucinations in RAG Applications Using an LLM's Internal States [0.5573267589690007]
  We focus on hallucinations involving information not used in training, which we determine by using recency to ensure the information emerged after a cut-off date.<n>This study investigates these hallucinations by detecting them at sentence level using different internal states of various language models.<n>Our results show that IAVs detect hallucinations as effectively as CEVs and reveal that answerable and unanswerable prompts are encoded differently as separate classifiers.
  arXiv  Detail & Related papers  (2024-12-22T15:08:24Z)
• Fine-grained Hallucination Detection and Editing for Language Models [109.56911670376932]
  Large language models (LMs) are prone to generate factual errors, which are often called hallucinations. We introduce a comprehensive taxonomy of hallucinations and argue that hallucinations manifest in diverse forms. We propose a novel task of automatic fine-grained hallucination detection and construct a new evaluation benchmark, FavaBench.
  arXiv  Detail & Related papers  (2024-01-12T19:02:48Z)
• "Knowing When You Don't Know": A Multilingual Relevance Assessment Dataset for Robust Retrieval-Augmented Generation [90.09260023184932]
  Retrieval-Augmented Generation (RAG) grounds Large Language Model (LLM) output by leveraging external knowledge sources to reduce factual hallucinations. NoMIRACL is a human-annotated dataset for evaluating LLM robustness in RAG across 18 typologically diverse languages. We measure relevance assessment using: (i) hallucination rate, measuring model tendency to hallucinate, when the answer is not present in passages in the non-relevant subset, and (ii) error rate, measuring model inaccuracy to recognize relevant passages in the relevant subset.
  arXiv  Detail & Related papers  (2023-12-18T17:18:04Z)
• HalluciDoctor: Mitigating Hallucinatory Toxicity in Visual Instruction Data [102.56792377624927]
  hallucinations inherent in machine-generated data remain under-explored. We present a novel hallucination detection and elimination framework, HalluciDoctor, based on the cross-checking paradigm. Our method successfully mitigates 44.6% hallucinations relatively and maintains competitive performance compared to LLaVA.
  arXiv  Detail & Related papers  (2023-11-22T04:52:58Z)
• A Stitch in Time Saves Nine: Detecting and Mitigating Hallucinations of LLMs by Validating Low-Confidence Generation [76.34411067299331]
  Large language models often tend to 'hallucinate' which critically hampers their reliability. We propose an approach that actively detects and mitigates hallucinations during the generation process. We show that the proposed active detection and mitigation approach successfully reduces the hallucinations of the GPT-3.5 model from 47.5% to 14.5% on average.
  arXiv  Detail & Related papers  (2023-07-08T14:25:57Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.